A Dynamic Overload Control Method for a High-Speed ​​Blender Based on Multi-Source Signal Fusion

Abstract

This invention discloses a dynamic overload control method for a blender based on multi-source signal fusion, specifically relating to the field of control and regulation. The method includes: first, performing multi-source signal acquisition and feature enhancement processing, simultaneously acquiring four types of signals: electrical, motion, mechanical, and thermal. After preprocessing, relevant feature sets are extracted. Second, conducting multi-dimensional complex coupling analysis, constructing functions to quantify single-field risks based on load intensity, mechanical state, and thermal safety, and designing three types of coupling terms—load and mechanical, mechanical and thermal, and load and thermal—to characterize multi-field synergistic effects. Then, generating a comprehensive risk value and dynamic threshold, solving for the risk value and its rate of increase, and setting thresholds hierarchically based on mode characteristics and identifying dominant coupling terms. Finally, executing four-level progressive control, initiating early warning preparation, intervention to break blockages, emergency evacuation, and final recovery or shutdown actions based on the risk status. This invention uses physical drive as its core to achieve precise risk quantification and coupling-guided control, balancing equipment safety and efficiency.

Description

Technical Field

[0001] This invention relates to the field of control and regulation technology, and more specifically, to a method for dynamic overload control of a blender based on multi-source signal fusion. Background Technology

[0002] High-speed blenders have been widely used in homes, restaurants, and other settings due to their high-speed pulverizing advantage. However, during operation, the motor needs to drive the cutter shaft at high speed. The load on the cutter shaft is prone to fluctuations due to changes in the characteristics of the ingredients and the amount added. At the same time, the high power output of the motor is accompanied by heat accumulation, and overload is a key issue that needs to be addressed during the operation of the equipment.

[0003] Currently, overload protection for blenders primarily relies on single-parameter monitoring technology. The industry generally uses sensors to collect single physical signals, commonly monitoring parameters such as motor operating current, winding temperature, and casing temperature. After collection, real-time parameters are directly compared to preset fixed thresholds. When a monitored parameter exceeds the threshold range, the equipment control system triggers preset protection actions, mainly including cutting off the motor power for emergency shutdown or automatically reducing the motor speed to reduce the load, thus achieving basic equipment safety protection. However, in actual use, it still has some drawbacks, such as…

[0004] 1. Limitations of single-parameter monitoring, overlooking risks of multi-field coupling: Existing methods mostly monitor only a single parameter such as current or temperature, ignoring the synergistic effects between load, mechanical, and thermal fields. For example, when relying solely on current to determine overload, it cannot identify mechanical jamming caused by low current but high vibration or hidden heat accumulation caused by load and thermal coupling. The isolation of parameters can easily lead to missed detections, resulting in tool shaft wear or hidden damage to the motor.

[0005] 2. Poor adaptability of fixed threshold and high misjudgment rate: The threshold is mostly a fixed value preset by the factory and is not adjusted in combination with the inherent load characteristics of the cell wall breaking mode. In the high load cell wall breaking mode, the fixed threshold is easy to trigger false shutdown, while in the low load stirring mode, the protection is delayed due to the high threshold. It cannot adapt to the risk differences of different working conditions, and it is difficult to balance user experience and equipment safety.

[0006] 3. Rigid protection actions, excessively affecting the performance: After the protection is triggered, it often performs a single shutdown action and lacks a gradient control mechanism; when faced with mild jamming or other recoverable overloads, direct shutdown not only interrupts the cell wall breaking process, but also requires the user to manually clean and restart; and when the risk increases sharply, the single action cannot quickly cut off the risk transmission chain, which can easily expand the scope of the fault.

[0007] 4. Lack of risk path identification and highly blind control: The core transmission path of overload is not located, and the protection actions lack specificity. For example, even with the same thermal overload, it is impossible to distinguish whether it is caused by load-thermal coupling or mechanical-thermal coupling. A uniform heat dissipation action is used, which is inefficient in controlling thermal runaway caused by vibration and friction and makes it difficult to accurately cut off the source of risk. Summary of the Invention

[0008] In order to overcome the above-mentioned defects of the prior art, the present invention provides a dynamic overload control method for a blender based on multi-source signal fusion, which solves the problems mentioned in the background art through the following scheme.

[0009] To achieve the above objectives, the present invention provides the following technical solution: a dynamic overload control method for a blender based on multi-source signal fusion, comprising:

[0010] S1: Multi-source signal acquisition and feature enhancement processing: After the equipment is started, electrical signals, motion signals, mechanical signals and thermal signals are acquired simultaneously; the acquired signals are preprocessed respectively, and then mechanical feature sets, vibration feature sets and thermal feature sets are extracted. The validity of the feature sets is verified by physical consistency.

[0011] S2: Multi-dimensional complex coupling analysis: Based on the feature set, basic sub-functions are constructed, including load intensity sub-function, mechanical state sub-function, and thermal safety sub-function; based on the basic sub-functions, cross-coupling terms between load and machinery, machinery and heat, and load and heat are designed to quantify the collaborative risks between different physical fields;

[0012] S3: Comprehensive Risk Value Generation and Dynamic Threshold Determination: The comprehensive risk value and its growth rate are solved by fusing the basic sub-functions and coupling terms through first-order differential equations; baseline thresholds are set based on the standard load capacity and ultimate load capacity of the model, and graded thresholds are dynamically generated and dynamically corrected in conjunction with the risk growth rate; the dominant coupling terms and risk transmission paths are identified by comparing the values ​​of coupling terms.

[0013] S4: Four-level control decision-making and execution: Control is executed based on comprehensive risk value, growth rate and dominant coupling term: the early warning level initiates adaptive preparatory actions; the intervention level outputs differentiated torque pulses and ultrasonic vibration composite actions; the emergency level cuts off the risk transmission chain according to the dominant coupling term; and the parameter gradient recovery or electrical and mechanical dual safety shutdown is executed according to the risk relief situation.

[0014] Preferably, the synchronous acquisition includes:

[0015] Electrical signal acquisition: The motor current is converted into an analog voltage signal by a preset sensor, which is then sampled by the ADC1 channel of the main control chip and converted into the actual current; the driving voltage is proportionally divided by a voltage sensor, sampled by the ADC2 channel and converted into the actual voltage, and then the input power is calculated and the motor efficiency is output based on the current and efficiency curve.

[0016] Motion signal acquisition: The pulses of the rotating shaft magnet are detected by a preset sensor, counted by an external interrupt, and converted into rotational speed and angular velocity; the voltage output by the strain gauge sensor is sampled by the ADC3 channel and converted into torque, and then the instantaneous stress is calculated.

[0017] Mechanical signal acquisition: The raw vibration data in the direction of the tool axis is output through the accelerometer via the I²C interface. The vibration amplitude is calculated every 10ms, and a set of vibration data is buffered every 100ms for spectrum preprocessing.

[0018] Thermal signal acquisition: The ambient temperature is collected by a temperature sensor, sampled through the ADC4 channel via an NTC resistor, the winding temperature is converted, and the heat generation power and heat dissipation power are calculated. At the same time, the ambient temperature is collected.

[0019] Synchronous triggering: The corresponding signal acquisition is triggered by setting the timer to 1ms, 5ms and 100ms period respectively, with a synchronization error of less than 0.1ms.

[0020] Preferably, the feature set includes:

[0021] Mechanical characteristic set: including normalized instantaneous stress, stress cycle number and load mechanical energy increment; wherein, the normalized instantaneous stress is the ratio of instantaneous stress to the yield limit of the cutter shaft, and the instantaneous stress is calculated from the torque and the torque and stress coefficient; the stress cycle number is the cumulative count when the torque fluctuation exceeds 10% of the rated value; the load mechanical energy increment is the numerical integral result of the product of torque and angular velocity within 100ms;

[0022] Vibration characteristic set: including vibration kinetic energy, damping dissipation energy and dominant frequency offset rate; wherein, vibration kinetic energy is obtained by multiplying the product of the tool shaft mass and the square of the vibration amplitude by 0.5; damping dissipation energy is the product of vibration kinetic energy and damping coefficient; dominant frequency offset rate is the ratio of the absolute value of the difference between the real-time vibration dominant frequency and the design natural frequency to the design natural frequency;

[0023] Thermal characteristic set: includes net thermal power and cumulative thermal energy; where net thermal power is the difference between heat generation power and heat dissipation power, heat generation power is obtained by multiplying the difference between input power and motor efficiency, heat dissipation power is the product of heat dissipation coefficient and the difference between winding temperature and ambient temperature; cumulative thermal energy is the numerical integral result of net thermal power over the operating time.

[0024] Preferably, the construction of basic sub-functions based on feature sets includes: constructing load strength sub-functions, mechanical state sub-functions, and thermal safety sub-functions, specifically as follows:

[0025] Load strength sub-function: The mechanical stress damage model is adopted, which is constructed based on the mechanical feature set and integrates the instantaneous stress over-limit damage and long-term fatigue cumulative damage effect; the square term of normalized instantaneous stress reflects the influence of instantaneous stress, and the ratio of stress cycle number to material fatigue life reflects fatigue damage. The weights of the two terms are adjusted by stress damage coefficient, and the results are mapped to the [0, 1] interval by exponential function.

[0026] Mechanical state sub-function: The vibration energy dissipation model is adopted and constructed based on the vibration feature set to reflect the synergistic effect of vibration energy accumulation and mode shift. The vibration energy dissipation state is reflected by the ratio of vibration kinetic energy to damping dissipated energy. The difference between the dominant frequency shift rate and the dominant frequency shift threshold is modulated by the modal sensitivity coefficient, and the influence weight of mode shift on risk is adjusted by an exponential function. The two are combined to calculate the sub-function value.

[0027] Thermal safety sub-function: It adopts a thermal balance model and is constructed based on thermal feature set to reflect the degree of thermodynamic energy imbalance. The cumulative thermal energy is used as the numerator, and the product of the motor heat capacity and the difference between the maximum allowable temperature and the ambient temperature is used as the denominator of the safe heat capacity. The ratio of the two quantifies the thermal risk level.

[0028] Preferably, the cross-coupling terms include: load-mechanical coupling terms, mechanical-thermal coupling terms, and load-thermal coupling terms, specifically:

[0029] Load and mechanical coupling term: Based on the mechanical and vibration feature set, the ratio of the vibration energy increment to the load mechanical energy increment within 100ms is calculated, and then multiplied by the normalized instantaneous stress modulated by the hyperbolic tangent function to quantify the conversion loss of mechanical energy to vibration energy.

[0030] Mechanical and thermal coupling term: Based on the vibration and thermal feature set, the ratio of the product of vibration kinetic energy and real-time main frequency to the cumulative load mechanical energy is calculated, and then multiplied by the temperature difference ratio modulated by the exponential function to quantify the conversion loss of vibration energy to thermal energy.

[0031] Load and thermal coupling term: Based on the mechanical, electrical and thermal feature set, calculate the product of the difference between 1 and motor efficiency and the ratio of real-time power to rated power, and then multiply it by a term containing cumulative energy correction to quantify the direct conversion of load power into heat energy.

[0032] Preferably, the process of solving for the comprehensive risk value and its growth rate includes:

[0033] Construct a first-order differential equation: the sum of the sum of the corresponding coupling terms and the risk growth coefficient of each basic sub-function is the risk increment, and the product of the comprehensive risk value and the decay coefficient is subtracted to obtain the comprehensive risk growth rate;

[0034] Numerical solution: The Euler method is used to solve the differential equation with a step size of 10ms. The current value is calculated by combining the comprehensive risk value of the previous time step. The value range is [0, 1].

[0035] Constraints are imposed: if the thermal safety subfunction is greater than or equal to 0.9, the growth rate is forcibly set to be greater than or equal to 0.05 / ms; if the load and mechanical coupling term is greater than or equal to 0.7, the corresponding risk growth coefficient is increased to 0.03.

[0036] Preferably, the dynamic generation of the hierarchical threshold includes:

[0037] Establish the baseline threshold: Subtract 0.1 times the ratio of the standard load capacity to the device's ultimate load capacity from 0.7 to obtain the baseline threshold;

[0038] The thresholds are divided into levels: an early warning threshold, an intervention threshold, an emergency threshold, and a fallback threshold are generated based on the baseline threshold. The early warning threshold is 0.8 times the baseline threshold, the intervention threshold is 0.9 times the baseline threshold, the emergency threshold is 1.05 times the baseline threshold, and the fallback threshold is equal to the baseline threshold.

[0039] Dynamic adjustment: If the overall risk growth rate is greater than or equal to 0.015 / ms, the tiered threshold is lowered by 10%; if the growth rate is less than or equal to 0.005 / ms, the threshold remains unchanged.

[0040] Preferably, the identification of the dominant coupling term includes:

[0041] Determine the objects to be compared: Select the calculated values ​​of load and mechanical coupling, mechanical and thermal coupling, and load and thermal coupling; compare the magnitudes of the three values ​​and take the coupling term corresponding to the maximum value as the dominant coupling term; if the difference between any two coupling terms is less than or equal to 0.1, it is determined to be a dual dominant coupling, and the risk transmission path corresponding to both types of coupling must be taken into account.

[0042] Preferably, the execution control includes:

[0043] Early warning level control: When the comprehensive risk value is between the early warning and intervention thresholds, the growth rate is greater than or equal to 0.01 / ms and the dominant coupling term is greater than or equal to 0.3, the corresponding preparatory action is initiated, prompting the optimization of the operating condition and continuous monitoring.

[0044] Intervention-level control: When the comprehensive risk value is between the intervention and emergency thresholds and lasts for two consecutive cycles without the warning being lifted, a differentiated torque pulse and ultrasonic vibration composite action is output to monitor the risk and the reduction of the coupling term;

[0045] Emergency-level control: When the first-level control fails to meet the standard, the risk is greater than or equal to the emergency threshold, or the growth rate is greater than or equal to 0.02 / ms, the risk chain is cut off according to the dominant coupling term, and the intensity of the action is dynamically adjusted every 10ms.

[0046] Final decision: If the recovery condition is met, the gradient callback parameters will be applied and the auxiliary system will be shut down after a delay; if the shutdown condition is met, the power supply will be cut off, mechanical protection will be activated, and alarm diagnosis will be initiated.

[0047] The technical effects and advantages of this invention are as follows:

[0048] 1. Multi-source fusion and coupling analysis for more comprehensive and accurate risk identification: By simultaneously collecting four types of signals—electrical, motion, mechanical, and thermal—and extracting multi-dimensional feature sets through preprocessing, the system then constructs cross-coupling terms of load and mechanical, mechanical and thermal, and load and thermal to quantify the inter-field synergistic effects. Compared to traditional single-parameter monitoring, it can fully cover the entire link of load, vibration, and thermal risks, capture hidden overload hazards from the mechanism level, significantly reduce the probability of missed detection, and improve the completeness and accuracy of risk identification.

[0049] 2. Dynamic threshold adaptation for enhanced working condition matching: Baseline thresholds can be set based on standard load patterns, and the tiered thresholds are dynamically adjusted according to the overall risk increase. The baseline is automatically adjusted to adapt to load characteristics under different operating modes, lowering the threshold to respond proactively when risk surges and maintaining threshold stability when risk is moderate. This solves the problems of misjudgment and lag associated with traditional fixed thresholds, achieving accurate risk assessment across all working conditions and adapting to the complex operational needs of blenders.

[0050] 3. Four-level progressive control to balance safety and user experience: Construct a four-level control system of early warning, intervention, emergency, and final decision-making. For mild risks, initiate pre-preparation; for moderate risks, implement de-blocking and evacuation; and for severe risks, implement mandatory control. Avoid process interruptions caused by traditional single shutdowns. While quickly containing risks and ensuring equipment safety, maximize the preservation of cell wall breaking effect, and balance safety and continuity of use.

[0051] 4. Coupled-term-oriented regulation for more targeted intervention: By comparing the values ​​of three types of coupled terms, the dominant risk transmission path is identified, and targeted regulation strategies are formulated. If load and mechanical coupling dominate, the load is reduced and vibration is suppressed; if load and thermal coupling dominate, power is reduced and heat dissipation is enhanced; if mechanical and thermal coupling dominate, friction is reduced and directional heat dissipation is achieved. This solves the problem of blind regulation in traditional methods and significantly improves the efficiency of overload intervention. Attached Figure Description

[0052] Figure 1 This is a schematic diagram of the overall structure of the present invention.

[0053] Figure 2 This is a schematic diagram of the S1 process of the present invention.

[0054] Figure 3 This is a schematic diagram of the S2 process of the present invention.

[0055] Figure 4 This is a schematic diagram of the S3 process of the present invention. Detailed Implementation

[0056] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0057] refer to Figures 1-4 The method for dynamic overload control of a blender based on multi-source signal fusion, as shown, includes:

[0058] S1: Multi-source signal acquisition and feature enhancement processing

[0059] Through a closed-loop process of pre-calibration, synchronous data acquisition, noise reduction, feature extraction, and quality verification, the physical signals of the blender during operation are transformed into a quantifiable and highly recognizable multi-dimensional feature system. This solves the problems of high noise, single dimension, and lack of related information in the original signals, providing accurate, complete, and relevant data input for the complex coupling and fusion of S2. The core logic is based on the principle of strong binding between signals, features, and physical meaning, ensuring that each type of feature can reflect the essential changes in equipment load, mechanical or thermal state. The specific steps are as follows:

[0060] S101: Multi-source signal synchronous acquisition: After the device starts up, it reads the mode code, loads the rated parameters, initializes the sensor and starts acquisition, and the status light stays on.

[0061] A1: Electrical signal acquisition:

[0062] Current Acquisition: The motor current is converted into a 0-3.3V analog voltage signal by a preset Hall current sensor, sampled by the ADC1 channel of the main control chip, and then analyzed using the formula... Convert to actual current; where For analog voltage signals, The midpoint voltage, I represents the sensor sensitivity, and I represents the actual current.

[0063] Voltage Acquisition: Based on a preset voltage sensor, the driving voltage is divided into 1 / 10 segments, sampled by the ADC2 channel, and then analyzed using the formula... Convert to actual voltage, where Where U is the voltage divider voltage, and U is the actual voltage.

[0064] Derivative calculation: Input power The average value is taken every 10 seconds as the effective power; then, the motor efficiency is output based on the real-time current I by fitting the current and efficiency curve. ;

[0065] A2: Motion signal acquisition:

[0066] Speed ​​acquisition: The pulses of the shaft magnet are detected by a Hall effect speed sensor, and the pulse count is recorded every 5ms via the external interrupt EXTI0 of the main control chip. Through formula Calculate the rotational speed (6-pole magnet, 6 pulses per revolution, 10ms converted to 100 cycles), and calculate the angular velocity: ;

[0067] Torque acquisition: The strain gauge sensor, integrated into the transmission coupling, outputs voltage in the stress concentration area of ​​the strain gauge blade shaft. After sampling by ADC3, the torque is obtained through conversion. Then calculate the instantaneous stress. ,in The torque-stress coefficient is collected every 1 ms;

[0068] A3: Mechanical signal acquisition:

[0069] Vibration amplitude acquisition: The original vibration data of the tool axis direction is output through the I²C interface via the MEMS accelerometer. After being read by the main control chip, the vibration amplitude A within 10ms is calculated, which is 1 / 2 of the difference between the peak value and the valley value.

[0070] Vibration spectrum preprocessing: Buffer a set of vibration data every 100ms;

[0071] A4: Thermal signal acquisition:

[0072] Temperature acquisition: The real-time ambient temperature is acquired through a preset temperature sensor and recorded as follows. The real-time temperature t of the winding coil is obtained by sampling the NTC resistor with temperature using ADC4 and converting it using the Steinhart-Hart formula.

[0073] Heat power calculation: Heat production power: Heat dissipation power: , This is the heat dissipation coefficient; updated every 100ms, in W.

[0074] A5: Baseline parameter calibration:

[0075] Tool Shaft Yield Limit The critical stress at which the cutter shaft material undergoes plastic deformation is obtained based on tensile tests of the cutter shaft specimens;

[0076] The mass m of the cutter shaft and the mass of the rigid body participating in the vibration are obtained based on weighing using an electronic balance.

[0077] The heat capacity of the motor, C, is the temperature rise of the motor after absorbing 1J of heat, obtained based on an adiabatic temperature rise experiment.

[0078] Design natural frequency The dominant frequency of undamped vibration in a mechanical system is determined based on modal excitation tests.

[0079] Damping coefficient c, the proportion of vibration energy dissipation, the rotation of the cutter shaft to 3000 r / min followed by release, and the amplitude decay curve over time are recorded. ,in The logarithmic decay rate;

[0080] Torque-stress coefficient The stress generated by the unit torque on the cutter shaft is determined by attaching four strain gauges to the middle section of the cutter shaft to form a full-bridge circuit. A gradient torque of 0-50 N·m is applied, and the slope of the stress-torque curve is fitted.

[0081] Heat dissipation coefficient The heat dissipation power of the motor when the temperature difference between the motor and the environment is 1℃ was obtained through a steady-state heat dissipation experiment.

[0082] It should be further noted that the main clock is 84MHz, timer TIM1 is configured with a 1ms period, TIM2 is configured with a 5ms period, and TIM3 is configured with a 100ms period.

[0083] Synchronization logic: When TIM1 is triggered, current, torque, and vibration are collected in parallel; when TIM2 is triggered, voltage and speed are collected in addition; when TIM3 is triggered, winding temperature and ambient temperature are collected in addition, with a synchronization error of less than 0.1ms.

[0084] The collected data were filtered and outlier removed before being normalized. The current and torque data were processed using Kalman filtering, and the vibration amplitude was filtered using wavelet thresholding. The temperature was calculated as an exponentially weighted moving average based on the rate of temperature change, and the rotational speed was filtered using median filtering.

[0085] S102: Physics-driven feature extraction: Based on preprocessed physical quantities, feature parameters are extracted and a feature set is constructed;

[0086] The set of mechanical characteristics includes: normalized instantaneous stress. : The stress cycle count N is incremented by one whenever the torque fluctuation exceeds 10%; the increase in load mechanical energy. , Numerical integration was performed using the trapezoidal rule with a step size of 1 ms.

[0087] Vibration characteristic set, including: vibration kinetic energy , Damping dissipates energy , ; frequency offset , , For real-time vibration dominant frequency, =500Hz;

[0088] Thermal characteristic set, including: net thermal power , Net accumulated heat energy , , For runtime, use the trapezoidal rule for numerical integration;

[0089] S103: Feature Quality Verification:

[0090] Mechanical characteristics: When it increases, The stress must be increased synchronously, and the difference in the increase deviation rate must be less than 15%; otherwise, it is judged as an abnormal stress calculation.

[0091] Vibration characteristics: When it increases, If it exceeds 0.15, then It must be greater than 1.5, otherwise the vibration signal is abnormal;

[0092] Thermal characteristics: When the temperature is increased, the winding temperature t must rise synchronously, with a temperature rise rate ≥ 0.05℃ / s; otherwise, it is judged as an abnormality in the thermal power calculation.

[0093] S2: Multidimensional Complex Coupling Analysis

[0094] Based on the mechanical, vibration, and thermophysical characteristics extracted by S1, a two-level analysis system of single-field risk quantification and multi-field coupling correlation is constructed. This system transforms the physical phenomena of equipment operation into quantifiable risk parameters, providing mechanistic input for comprehensive risk generation. This overcomes the limitations of traditional analysis that isolates and assesses single-dimensional risk. The specific analysis is as follows:

[0095] S201: Construction of Basic Sub-functions: Based on physical laws, this method decomposes single-dimensional risks by constructing sub-functions for three core physical fields: load impact, mechanical vibration, and thermal accumulation. Using fundamental functions from materials mechanics, vibration theory, and thermodynamics, the characteristic parameters output by S1 are transformed into quantitative indicators reflecting the intensity of single-field risks. Each sub-function integrates instantaneous and cumulative effects, as detailed below:

[0096] B1: Load Strength Subfunction —Mechanical stress damage model:

[0097] The damage to equipment caused by load consists of two parts: plastic deformation caused by instantaneous stress exceeding the material's yield limit, and fatigue damage caused by long-term alternating stress. This conforms to the stress-life curve law in mechanics of materials, and its specific mathematical function is as follows:

[0098]

[0099] in, The stress damage coefficient is the weight of the stress on equipment damage. It is calibrated by accelerated life test of the tool shaft material and is set to 0.8. The larger the value, the faster the damage accumulates under the same stress. The fatigue life of the shaft material is obtained based on the factory calibration.

[0100] B2: Mechanical State Subfunction —Vibrational energy dissipation model:

[0101] When a mechanical system is operating normally, vibration energy is dissipated through the damping structure. When abnormalities such as jamming or bearing wear occur, the damping dissipation capacity decreases, vibration energy accumulates, and the system's natural frequency deviates from the design mode. Both factors work together to reflect the mechanical health status, and their specific mathematical functions are as follows:

[0102]

[0103] in, The damped dissipated energy, in J, is the portion of the vibrational energy consumed by the damped structure. The calculation formula is... , The modal sensitivity coefficient reflects the degree of impact of the dominant frequency shift on mechanical risk. It is set to 20. The larger the value, the steeper the risk increases during modal shift. The main frequency offset threshold is a critical value for judging modal anomalies, and its value is 0.15.

[0104] B3: Thermal safety subfunction —Thermal equilibrium model:

[0105] The thermal risk of an electric motor depends on the balance between heat generation and heat dissipation. When the heat generation power exceeds the heat dissipation power, heat energy accumulates, and the temperature rises. The thermal risk is the ratio of net accumulated heat energy to the motor's safe heat capacity, directly reflecting the degree of thermodynamic energy imbalance. Its specific mathematical function is as follows:

[0106]

[0107] in, This refers to the maximum allowable temperature of the motor.

[0108] S202: Cross-coupling term design: Based on the multi-dimensional risk superposition of energy conversion, and the physical conduction chain of load mechanical energy, mechanical vibration energy, and heat dissipation, three types of coupling terms are designed to quantify the mutual reinforcement effect between different physical fields, avoiding the omission of collaborative risks in single-dimensional analysis. All coupling terms are based on energy conversion efficiency, and the weights are adjusted in combination with key state parameters, as detailed below:

[0109] C1: Load and mechanical coupling term —Loss in the conversion of mechanical energy and vibrational energy:

[0110] Under high load, part of the input mechanical energy is used for cell breaking as effective work, and part is converted into mechanical vibration energy as harmful work. The coupling term reflects the proportion of harmful work and is strengthened with the increase of instantaneous stress. The higher the stress, the higher the efficiency of mechanical energy to vibration energy conversion, forming a positive feedback loop where increased load leads to increased vibration. The specific mathematical function is as follows:

[0111]

[0112] in, This represents the vibration energy increment, measured in J, and is the difference in vibration kinetic energy between the current 100ms and the previous 100ms. This reflects the change in vibration energy caused by load changes; For the stress modulation term, the hyperbolic tangent function maps the stress normalization value to [0, 1];

[0113] C2: Mechanical and thermal coupling term —Loss in the conversion of vibrational energy and thermal energy:

[0114] When mechanical vibration intensifies, friction between components increases, raising the proportion of vibration energy converted into heat. Simultaneously, vibration may lead to poor fit of the heat dissipation structure, reducing heat dissipation efficiency. Increased temperature softens the lubricating grease, further increasing friction, creating a coupling effect where enhanced vibration leads to accelerated heat accumulation. The specific mathematical function is as follows: ;

[0115] in, It represents the cumulative load mechanical energy, measured in J, and is the total load energy from equipment startup to the present, reflecting the total working intensity of the mechanical system. The temperature modulation term indicates that the exponential function grows non-linearly with increasing temperature;

[0116] C3: Load and thermal coupling term —Direct conversion of load power and heat energy:

[0117] When the load increases, the motor output torque increases, the current increases, and losses such as copper loss and iron loss increase, resulting in decreased motor efficiency and a higher proportion of input power converted into heat energy. Simultaneously, prolonged high-power operation leads to continuous heat accumulation, which can cause thermal overload even if the instantaneous load does not reach its limit. The specific mathematical function is:

[0118]

[0119] in, Rated power, Baseline runtime;

[0120] S3: Comprehensive Risk Value Generation and Dynamic Threshold Determination

[0121] Based on the subsystem risk value and multi-field coupling terms output by S2, a comprehensive risk value is generated by fusing the physical-driven dynamic state equations. This is combined with the risk evolution trend and operating condition characteristics to generate an adaptability threshold, achieving a precise transformation from multi-dimensional mechanism parameters to single-value decision indicators. This provides a quantitative and executable basis for subsequent graded control. The specific steps are as follows:

[0122] S301: Comprehensive Risk Value Generation: Based on multi-parameter fusion of risk evolution laws, the dynamic process of risk accumulation and attenuation is described through first-order differential equations. Subsystem risks and coupling effects are used as sources of risk increment. An attenuation term is introduced to simulate the natural dissipation of risk when there are no equipment anomalies. The real-time comprehensive risk value is obtained through numerical solution, reflecting the current risk intensity and changing trend. The specific mathematical function of the comprehensive risk growth rate is as follows:

[0123]

[0124] in, , , The risk growth coefficients reflect the contribution weight and growth rate characteristics of each subsystem's risk to the overall risk, with values ​​of 0.02, 0.015, and 0.01, respectively. This is the risk attenuation coefficient, reflecting the natural dissipation rate of risk when there are no anomalies; its value is 0.005. The overall risk growth rate; To calculate the comprehensive risk value, the Euler method is used for numerical solution. The solution step size is consistent with the data acquisition and analysis cycle to ensure data synchronization. The comprehensive risk value calculation function is as follows:

[0125]

[0126] Its value range is [0, 1]. Then force setting ,like ,but Temporarily increased to 0.03 / ms;

[0127] S302: Dynamic Threshold Generation: Based on the inherent energy characteristics of different operating modes, a baseline threshold is set, and then dynamically adjusted in conjunction with real-time risk increase. This ensures that the threshold can adapt to different operating conditions such as high load and low load, and can also cope with sudden scenarios such as a sudden increase in risk, reducing misjudgment and missed judgment. The specific analysis is as follows:

[0128] Baseline threshold: , For standard load capacity, This is the equipment's maximum load capacity;

[0129] Grading threshold definition: Based on baseline threshold, four levels of judgment criteria are defined.

[0130] Warning threshold : The risk has entered a period of concern, and preparatory work needs to be initiated.

[0131] Intervention threshold : When the risk reaches the intervention threshold, proactive measures such as breaking the blockade or diverting the flow of information are required.

[0132] Emergency threshold : The risk is nearing its limit, and a forced load reduction is necessary.

[0133] bottom line threshold The risk has reached the physical limit, and the machine must be shut down immediately.

[0134] Risk growth rate correction: The correction rule is as follows: if the real-time risk growth rate is... If the risk increases by 15% per second, all risk thresholds will be lowered by 10%. If the growth rate is less than 0.005 / ms, it indicates that the risk is changing slowly and the threshold remains unchanged.

[0135] S303: Dominant Coupling Item Identification: Based on risk transmission path strategy adaptation, by comparing the numerical values ​​of the three types of coupling items, the dominant transmission path of the current risk is identified, providing a clear direction for subsequent coupling-oriented regulation, and differentiating measures are formulated for the dominant coupling item;

[0136] Recognition method: Calculate the output of S2 , , The value is used to determine the dominant coupling term. If the difference between two coupling terms is less than or equal to 0.1, it is determined to be a dual dominant coupling, and the regulation of both types of paths needs to be taken into account at the same time.

[0137] Correspondence between dominant coupling terms and risk transmission paths:

[0138] Increased load leads to enhanced vibration and mechanical damage; therefore, reduce the load and use ultrasonic decompression.

[0139] Increased load leads to decreased efficiency and accelerated heat accumulation, so power should be reduced and heat dissipation should be strengthened.

[0140] Increased vibration leads to increased friction and thermal runaway, so reduce the rotation speed and implement directional heat dissipation.

[0141] S4: Four-level control decision-making and execution based on dynamic risk

[0142] Based on the comprehensive risk value, risk growth rate, dynamic classification threshold and dominant coupling term output by S3, the system implements progressive control of early warning, barrier breaking, guidance and decision-making. The core objective is to maximize the barrier breaking effect while ensuring equipment safety. By dynamically adjusting the coupling term to adapt to the action and risk growth rate, the system solves the problems of rigid actions and high misjudgment rate in traditional protection.

[0143] S401: Core Decision-Making System: Construct a three-dimensional decision matrix of static risk value, dynamic growth rate, and coupling path, where: static value matches the control level, growth rate determines the urgency of response, coupling term locates the intervention target, and the three work together to ensure accurate, timely and appropriate actions;

[0144] Its decision-making dimensions and guiding role in regulation are as follows:

[0145] Static risk intensity: Determine the control level: early warning, intervention, emergency, and safety net;

[0146] Dynamic risk trends: Adjust action intensity and correct trigger delays;

[0147] Risk transmission path: dominant coupling factors, locating intervention targets;

[0148] Action effect feedback: Dynamically adjust action parameters to form a closed loop of execution, monitoring, and correction;

[0149] S402: Level 0: Early Warning: This level is for scenarios where risks are initially apparent but have not reached the intervention threshold. The core is to predict the direction of risk transmission and initiate preparatory actions, which neither interferes with user operations nor shortens the response time for possible subsequent adjustments. It is equivalent to the system's early warning buffer.

[0150] Triggering conditions: Triple verification to avoid false alarms; the following conditions must be met simultaneously:

[0151] (1) The overall risk level has entered the warning range;

[0152] (2) The risk is on the rise, excluding instantaneous fluctuations;

[0153] (3) The dominant coupling term is greater than or equal to 0.3, indicating a clear risk transmission path and non-random noise;

[0154] The above three points are collected continuously for three collection cycles to avoid signal glitches that could trigger the signal.

[0155] Execution actions: Preparatory actions initiated for the dominant coupling item:

[0156] The ultrasonic vibrator enters low-power standby mode, and the electromagnetic clutch is pre-magnetized.

[0157] The cooling fan is set to its lowest speed.

[0158] The grease-heated plate starts at low power, and the fan speed increases to 60% of the rated value.

[0159] OLED screen display: Under operating condition optimization, the indicator green light flashes slowly at 0.5Hz; the buzzer is silent; parameters are updated every 10ms. If the conditions are met for 5 consecutive cycles (50ms), the result will be satisfactory. If all coupling terms are less than 0.3, then the pre-preparation action is disabled and normal operation is resumed;

[0160] S403: Level 1: Adaptive Torque Pulse: This level is for scenarios where the risk reaches the intervention threshold but is not out of control. The core is to generate differentiated torque pulses and ultrasonic vibration composite actions based on coupling strength and stress magnitude. The instantaneous impact releases the stuck objects and avoids direct deceleration that would affect the cell wall breaking effect.

[0161] Triggering conditions: The following conditions must be met simultaneously:

[0162] (1) The overall risk has reached the intervention threshold;

[0163] (2) This state has lasted for two data collection cycles, the risk is stable, and it is not a sudden shock;

[0164] (3) The Level 0 warning has not been lifted.

[0165] Differentiated action execution:

[0166] like and For strong coupling and hard jamming, high-frequency ultrasonic resonance is used to release the jammed object and short-cycle reverse reversal is used to throw it away. The low degree of engagement reduces the impact transmission. The specific actions are as follows: voltage is increased by 15%, pulse duration is 180ms, ultrasonic frequency is 32kHz, forward and reverse rotation cycle is 1s forward and 0.2s reverse, and the clutch engagement degree is set to 60%.

[0167] like and For weak coupling and soft jamming, a long pulse torque is continuously applied and low-frequency ultrasonic softening is used. High engagement ensures torque transmission efficiency. The specific actions are as follows: voltage is increased by 13%, pulse duration is 220ms, ultrasonic frequency is 28kHz, forward and reverse rotation cycle is 1.2s forward and 0.3s reverse, and the clutch engagement degree is set to 80%.

[0168] High-frequency monitoring and effect assessment (100ms window period)

[0169] Immediately after the control measures are implemented, the system switches to high-frequency monitoring with a 2ms sampling period, focusing on tracking two core indicators:

[0170] Risk reduction: ,Target The risk has decreased significantly;

[0171] Coupling term reduction: ,Target The coupling chain is severed;

[0172] Judgment result:

[0173] If both parameters meet the target within 100ms, then proceed to the 500ms observation period; if no abnormalities are found, restore the normal parameters.

[0174] Failure to meet standards: If any indicator fails to meet the standards, the regulation will be immediately upgraded to the second level.

[0175] S404: Level 2: Coupling-guided load diversion: This level is designed for scenarios where Level 1 control fails or risks increase sharply. The core is to cut off the core link of risk transmission based on the dominant coupling item. For load and mechanical chain, the load is reduced and vibration is reduced. For load and thermal chain, the power is reduced and heat dissipation is strengthened. For mechanical and thermal chain, friction is reduced and directional heat dissipation is strengthened, so as to achieve precise risk reduction rather than blind shutdown.

[0176] Triggering condition: Meets any of the following conditions:

[0177] The 100ms monitoring for Level 1 control failed to meet the standard.

[0178] The overall risk has reached the emergency threshold;

[0179] This means a surge in risk growth;

[0180] Decoupling term guidance strategy:

[0181] Dominate, reduce load input, and weaken the transmission of vibration to the motor:

[0182] Load control: Stepped speed reduction, each step reducing the rated speed by 5% at 200ms intervals, eventually reducing to 55% of the initial speed;

[0183] Vibration suppression: The ultrasonic vibrator is switched to a strong vibration mode of 150ms on / 50ms off to directionally loosen the entangled material;

[0184] Mechanical isolation: The electromagnetic clutch engagement is dynamically maintained at 40% to 50%, absorbing vibration and shock through slip and reducing transmission to the motor spindle;

[0185] Dominant, reducing power loss and accelerating heat dissipation:

[0186] Power limiting: The input power is locked at 70% of the rated power via a PWM signal;

[0187] Enhanced heat dissipation: Fans at full speed, with added 2s on / 100ms off pulse-type power-off cooling;

[0188] Load optimization: Speed ​​reduced to 70% of initial speed;

[0189] Dominate, reduce mechanical friction, and enhance heat dissipation in key components:

[0190] Friction control: The rotation speed is reduced to 65% of the initial rotation speed, and the ultrasonic vibrator is switched to 20kHz low frequency mode;

[0191] Targeted cooling: Adjusting the angle of the fan shroud so that the airflow is directly directed at the motor bearing, increasing the bearing temperature reduction by 30%;

[0192] Enhanced lubrication: The power of the grease heating element has been increased to 5W;

[0193] Dynamic correction mechanism

[0194] Calculated every 10ms The rate of change and the rate of change of the dominant coupling term are adjusted in real time.

[0195] like Immediately increase the intensity of the corresponding movements, i.e., increase each movement by 5%;

[0196] like Forced full-speed cooling and reduced speed to 60%;

[0197] like Immediately reduce the speed by an additional 15% to prevent plastic deformation of the cutter shaft.

[0198] S405: Level 3: Final Decision: This level is the risk acceptance stage 300ms after the risk has been mitigated. The core is to execute gradient recovery or safe shutdown based on whether the risk has been eliminated.

[0199] To determine whether operation can be restored, four conditions must be met simultaneously:

[0200] (1) That is, the comprehensive risk returns to the safe range;

[0201] (2) All coupling terms are less than 0.3, meaning the potential conduction chain is completely broken;

[0202] (3) The motor temperature is less than 60℃, that is, the thermal safety margin is sufficient;

[0203] (4) Rotation speed fluctuation is less than 5%.

[0204] Action performed: Parameter gradient recovery: Restore the speed, voltage, and clutch engagement to the current mode baseline parameters at a rate of 5% / 100ms;

[0205] Auxiliary system delay: The ultrasonic vibrator and enhanced heat dissipation system will continue to run for 1 second before shutting down;

[0206] An emergency shutdown is determined if any of the following conditions are met:

[0207] (1) This means that the risk is still within the intervention range 300ms after the intervention.

[0208] (2) The motor temperature exceeds 110℃, which is close to the insulation limit of 120℃;

[0209] (3) The speed drops to 60% and .

[0210] Perform the following actions:

[0211] Electrical protection: The relay cuts off the main power supply, and the PWM module immediately returns to zero;

[0212] Mechanical protection: The electromagnetic clutch is fully disengaged, and the mechanical locking device extends its locking tongue;

[0213] Fault diagnosis: Retrieve historical data from S2 / S3 and locate the cause of the fault based on the dominant coupling term;

[0214] User alarm: OLED screen display malfunction, red light stays on and 5Hz continuous alarm sounds;

[0215] Secondly: The accompanying drawings of the embodiments disclosed in this invention only involve the structures involved in the embodiments disclosed in this invention. Other structures can refer to the general design. In the absence of conflict, the same embodiment and different embodiments of this invention can be combined with each other.

[0216] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for dynamic control of an overload of a cell wall breaking machine based on multi-source signal fusion, characterized in that, Comprise: S1: multi-source signal acquisition and feature enhancement processing: after the device starts, the electrical signal, motion signal, mechanical signal and thermal signal are synchronously collected; The collected signals are preprocessed, and then the mechanical feature set, vibration feature set and thermal feature set are extracted, and the effectiveness of the feature set is verified through physical consistency; S2: multi-dimensional complex coupling analysis: based on the feature set, the basic sub-function is constructed, including load intensity sub-function, mechanical state sub-function and thermal safety sub-function; Based on the basic sub-function, the cross-coupling terms of load and mechanical, mechanical and thermal, and load and thermal are designed, and the collaborative risk between different physical fields is quantified; The basic sub-function is constructed based on the feature set, specifically including: Load intensity sub-function: using a mechanical stress damage model, based on the mechanical feature set, the instantaneous stress over-limit damage and long-term fatigue cumulative damage effect are fused; the square term of normalized instantaneous stress is used to reflect the influence of instantaneous stress, and the ratio of stress cycle number to material fatigue life is used to reflect fatigue damage, the weight of the two is adjusted by stress damage coefficient, and the result is mapped to the interval [0, 1] by exponential function; Mechanical state sub-function: using a vibration energy dissipation model, based on the vibration feature set, the collaborative influence of vibration energy accumulation and modal shift is reflected; the ratio of vibration kinetic energy to damping dissipation energy reflects the vibration energy dissipation state, and the difference between the main frequency shift rate and the main frequency shift threshold is modulated by the modal sensitivity coefficient, then the influence weight of modal shift on risk is adjusted by exponential function, and the sub-function value is calculated by combining the two; Thermal safety sub-function: using a thermal balance model, based on the thermal feature set, the degree of thermodynamic energy imbalance is reflected; the cumulative thermal energy is the numerator, and the product of the motor heat capacity and the difference between the highest allowable temperature and the ambient temperature is the denominator, and the ratio of the two quantifies the thermal risk level; The cross-coupling term specifically includes: Load and mechanical coupling term: based on the mechanical and vibration feature sets, the ratio of vibration energy increment to load mechanical energy increment within 100ms is calculated, and then multiplied by the normalized instantaneous stress modulated by the hyperbolic tangent function, to quantify the mechanical energy conversion loss to vibration energy; Mechanical and thermal coupling term: based on the vibration and thermal feature sets, the ratio of the product of vibration kinetic energy and real-time main frequency to cumulative load mechanical energy is calculated, and then multiplied by the temperature difference ratio modulated by the exponential function, to quantify the vibration energy conversion loss to thermal energy; Load and thermal coupling term: based on the mechanical, electrical and thermal feature sets, the product of the difference between 1 and motor efficiency and the ratio of real-time power to rated power is calculated, and then multiplied by the term containing cumulative electrical energy correction, to quantify the direct conversion of load power to thermal energy; S3: comprehensive risk value generation and dynamic threshold determination: the basic sub-function and coupling term are fused by a first-order differential equation, and the comprehensive risk value and growth rate are solved; based on the mode standard load energy and the limit load energy, the baseline threshold is set, and the hierarchical threshold is dynamically generated, which is dynamically corrected combined with the risk growth rate; the dominant coupling term and risk transmission path are identified by comparing the coupling term values. S4: Fourth-level regulation decision and execution: based on the comprehensive risk value, the growth rate and the dominant coupling term, execute regulation: the early warning level starts the adaptive preparation action; the intervention level outputs the differentiated torque pulse and the ultrasonic vibration composite action; the emergency level cuts off the risk transmission chain according to the dominant coupling term; according to the risk removal condition, execute parameter gradient recovery or electrical and mechanical double safety shutdown.

2. The method according to claim 1, wherein the method is characterized in that: The synchronous acquisition comprises: Acquisition of electrical signals: convert motor current into analog voltage signals through a preset sensor, sample and convert actual current through the main control chip ADC1 channel; divide the driving voltage by a proportional voltage divider through a voltage sensor, sample and convert the actual voltage through the ADC2 channel, and then calculate the input power and the motor efficiency based on the current and efficiency curve; Acquisition of motion signals: detect the rotor shaft magnetic steel pulse through a preset sensor, count and convert the rotational speed and angular velocity through an external interrupt; output the voltage through a strain gauge sensor, sample and convert the torque through the ADC3 channel, and then calculate the instantaneous stress; Acquisition of mechanical signals: output the original data of the cutter shaft direction vibration through an accelerometer through an I2C interface, calculate the vibration amplitude within every 10 ms, and cache a group of vibration data every 100 ms for frequency spectrum preprocessing; Acquisition of thermal signals: collect the ambient temperature through a temperature sensor, sample through the ADC4 channel through an NTC resistor, convert the winding temperature, and then calculate the heat generation power and the heat dissipation power while collecting the ambient temperature; Synchronous triggering: set 1 ms, 5 ms and 100 ms periods through a timer to trigger corresponding signal acquisition respectively, and the synchronization error is less than 0.1 ms.

3. The method according to claim 1, wherein the method is characterized in that: The feature set comprises: Mechanical feature set: including normalized instantaneous stress, stress cycle number and load mechanical energy increment; wherein the normalized instantaneous stress is the ratio of the instantaneous stress to the yield limit of the cutter shaft, the instantaneous stress is calculated from the torque and the torque and stress coefficient; the stress cycle number is the cumulative count when the torque fluctuation exceeds 10% of the rated value; and the load mechanical energy increment is the numerical integration result of the product of the torque and the angular velocity within 100 ms; Vibration feature set: including vibration kinetic energy, damping dissipation energy and main frequency shift rate; wherein the vibration kinetic energy is obtained by multiplying the mass of the cutter shaft and the square of the vibration amplitude by 0.5; the damping dissipation energy is the product of the vibration kinetic energy and the damping coefficient; and the main frequency shift rate is the ratio of the absolute value of the difference between the real-time vibration main frequency and the designed natural frequency to the designed natural frequency; Thermal feature set: including net heat power and cumulative heat energy; wherein the net heat power is the difference between the heat generation power and the heat dissipation power, the heat generation power is obtained by the product of the difference between the input power and the motor efficiency, and the heat dissipation power is the product of the heat dissipation coefficient and the difference between the winding temperature and the ambient temperature; and the cumulative heat energy is the numerical integration result of the net heat power within the running time.

4. The method according to claim 1, wherein the method is characterized in that: The solving of the comprehensive risk value and the growth rate comprises: Constructing a first-order differential equation: multiplying each basic sub-function by the sum of the corresponding coupling term and the sum of the risk growth coefficient to obtain the risk increment, subtracting the product of the comprehensive risk value and the attenuation coefficient, and the result is the comprehensive risk growth rate; Numerical solution: solve the differential equation by using the Euler method with a step size of 10 ms, and calculate the current value based on the previous time comprehensive risk value, which is in the range of [0, 1]; Constraint application: if the thermal safety sub-function is greater than or equal to 0.9, the acceleration is forced to be greater than or equal to 0.05 / ms; if the load and mechanical coupling term is greater than or equal to 0.7, the corresponding risk growth coefficient is increased by 0.

03.

5. The method according to claim 1, wherein the method is characterized in that: The dynamic generation of the hierarchical threshold value comprises: Baseline threshold value construction: the baseline threshold value is obtained by subtracting 0.1 times the ratio of the mode standard load capacity to the equipment limit load capacity from 0.7; Hierarchical threshold value division: the early warning threshold value, the intervention threshold value, the emergency threshold value and the bottom threshold value are generated based on the baseline threshold value, wherein the early warning threshold value is 0.8 times the baseline threshold value, the intervention threshold value is 0.9 times the baseline threshold value, the emergency threshold value is 1.05 times the baseline threshold value, and the bottom threshold value is equal to the baseline threshold value; Dynamic correction: if the comprehensive risk growth rate is greater than or equal to 0.015 / ms, the hierarchical threshold value is decreased by 10%; if the growth rate is less than or equal to 0.005 / ms, the threshold value remains unchanged.

6. The method according to claim 1, wherein the method is characterized in that: The dominant coupling term identification comprises: Determination of the object to be compared: the calculated values of the load and mechanical coupling term, the mechanical and thermal coupling term, and the load and thermal coupling term are selected; the maximum value of the three values is taken as the dominant coupling term; if the difference between the values of any two coupling terms is less than or equal to 0.1, it is determined that there are two dominant couplings, and both coupling paths need to be considered.

7. The method according to claim 1, wherein the method is characterized in that: The execution control comprises: Early warning level control: when the comprehensive risk value is between the early warning and intervention threshold values, the growth rate is greater than or equal to 0.01 / ms, and the dominant coupling term is greater than or equal to 0.3, the corresponding pre-preparation action is started, and the working condition optimization is prompted and continuously monitored; Intervention level control: when the comprehensive risk value is between the intervention and emergency threshold values and lasts for 2 cycles, and the early warning is not removed, the differential torque pulse and ultrasonic vibration composite action are output, and the risk and coupling term decrease are monitored; Emergency level control: when the first level control is not up to standard, the risk is greater than or equal to the emergency threshold value, or the growth rate is greater than or equal to 0.02 / ms, the risk chain is cut off according to the dominant coupling term, and the action intensity is dynamically corrected every 10ms; Final decision: if the recovery condition is reached, the parameters are gradually adjusted back, and the auxiliary system is delayed to be closed; if the shutdown condition is reached, the power supply is cut off, the mechanical protection is performed, and the alarm diagnosis is performed.

Images